Luth-2 Released: Sets New SOTA for French Small Language Models
Unusual_Shoe2671 · reddit · 2026-08-11
kurakurai has released Luth-2-0.8B and Luth-2-2B, two new non-reasoning French language models that achieve state-of-the-art results in their size class. They outperform models 3-4 times their size on various French benchmarks like Multi-IF, MGSM-Rev2, and Math-500.
Key Technical Improvements:
- Introduces a new 3B-token SFT mixture covering math, knowledge, code, tool calling, and multi-turn dialogue.
- Utilizes reinforcement learning via expert specializations and multi-domain on-policy distillation (MOPD).
- Switches to Qwen3.5 as the base architecture, which proved significantly more receptive to post-training.
Both models are lightweight enough for on-device local use, and their weights, datasets, and code are open-sourced on Hugging Face.
More from Models
- UCL Researchers Use RL to Stop LLMs from Lying About Their Hidden Reasoning — alex_verem · 2026-08-11
- Falcon-Perception: A 0.6B Model for Generating Labels for Object Detection and Segmentation — vanstriendaniel · 2026-08-11
- Claude Outputs Now Include Text Watermarks, Sparking Removal Discussions — Franck_Dernoncourt · 2026-08-11
- Predictions: Major Update by Late Sept, Next Paper to Focus on Agents — teortaxesTex · 2026-08-11
- Anthropic to embed invisible watermarks in all Claude text outputs globally — The Decoder · 2026-08-11
- Rumor: DeepSeek Holding onto V4 Pro GA Release — dejavucoder · 2026-08-11